Many organizations maintain documents that include information, such as of a medical or financial nature, that is deemed sensitive and whose exposure to individuals is restricted by company policy, or even by law. Such organizations may employ document redaction software that removes sensitive information from documents before they are shown to individuals requesting access to such documents. Document redaction software typically removes sensitive information from documents automatically according to a set of predefined rules that identify sensitive information within documents by using dictionaries of sensitive words, such as of given names and surnames, or by using visual characteristics, such as where a rule dictates that all 16-digit numbers that appear within a document be removed on the assumption that they are credit card numbers.